Microsoft 365 Copilot: What we know so far
Five months ago, Microsoft announced the launch of the very aptly named M365 “Copilot”, an AI-driven tool that leverages large language models (LLMs) to serve as your digital personal assistant.
Described by Microsoft CEO Satya Nadella as “the next platform shift”, Copilot combines natural language recognition with a reasoning engine. The result: M365 can translate natural language requests into detailed actions which it then completes on your behalf. The overall impact to productivity? As outlandish as it sounds, Microsoft are predicting a 7 – 10% increase in GDP from countries (and organisations) who can effectively leverage this generation of AI technologies.
Unsurprisingly this has generated a lot of attention from Microsoft customers and partners across the world. Unusually coy, Microsoft haven’t yet announced a release date for the service (currently in private trial with 600 hand-selected organisations) but have announced a price: $30 per user, per month.
Couple of points to reflect on here:
- $30 per user per month is a substantial uplift (this is in addition to a minimum E3 / ideally E5 license). With this being more than expected, there will likely be significant pressure on organisations to demonstrate rapid time-to-value here.
- However, if the 7 – 10% efficiency improvement figure is achieved (and that is a big if)thenthis equates to an absolute bargain – productivity vs cost breaks even for anyone earning over $300 – $429 per month, anything beyond that is upside.
Clearly high stakes with high rewards available for organisations who get it right. As with all innovation, this will likely lead to an “arms race” with organisations able to capitalise accelerating their own growth at the expense of organisations who either get it wrong or don’t adopt at all. Thorough planning is going to be essential, some things you may want to consider:
- Have you fully migrated all relevant data to the Cloud? Tools like OneDrive, SharePoint, OneNote, Teams etc. all come into the fray here, Copilot is only as good as the data it has access to.
- Is your data secure with robust governance processes? In the past data has been able to “hide” in long forgotten corners of your estate, protected by lack of awareness rather than any formal security policies. Do you want to shine a floodlight on that legacy data for all your employees to access?
- Roll-out strategy – will all users benefit equally from this? This is where personas come in, how can you manage your roll-out to deliver a steady ROI that enables you to scale incrementally?
- End user adoption – as anyone who’s used ChatGPT knows, asking the right questions is key. This practice (known as “prompt engineering”) will likely be the differentiator between adoption success or failure. How do you upskill your team to make the most of this new tool?
This is far from an exhaustive list but hopefully gives some form of steer as to key topics for consideration. I’ll be exploring these in more depth in future blogs. For now, if you want to be kept up to date on developments as they emerge, please sign up to our monthly Copilot webinar via the link here.